Can historical controls be used in current clinical trials in osteosarcoma? Analysis of prognostic factors

A comparison with respect to possible prognostic factors has been made between a contemporary group of 44 patients with osteosarcoma in Sweden, who received treatment since 1971, and a historical group of 35 such patients treated prior to 1972. Only patients with no evidence of metastases on admission were included in the material. The case histories were reviewed by a team of pathologists and clinicians with experience of bone tumors, each case history was analyzed in detail and re-evaluated. The female/male ratio was lower, but not significantly so, for the contemporary group, 1:2.4, than for the historical group, 1:1.7. The mean age was similar in the two groups (18 and 17 years, resp.) Pain was the most common symptom in both groups, it was recorded in 86 and 100% in the contemporary and historical groups, respectively. Although the patients of the historical group generally presented more symptoms, the durations were similar in the two groups (3 months). The site patterns differed slightly, although not significantly so, in the two groups, there was a higher ratio of frequency of femoral and lower leg tumors for the historical group (54 and 32%, resp.) than for the contemporary group (36 and 45 %, resp.). The variability in tumor size was greater for the historical group, as was the tumor diameter (13 and 9 cm, resp.). The historical group had a greater proportion of osteoblastic tumors (83 and 63%, resp.) and these tended to be more malignant (Grade IV:80 and 68%, resp.). The study disclosed differences between the contemporary and the historical groups in respect to various factors that previous studies indicated might have a bearing on the prognosis in osteosarcoma. On the whole the prognosis would be expected to be less favorable for the historical group and there would seem to be a definite risk in using this group as a control group to the adjuvant interferon treated patients. Similar reviews of other non-randomized current clinical series would be a valuable aid in determining the efficacy of adjuvant therapy.